Wedge

I'm working on it because I learned a very hard and very valuable lesson during SD's development. Namely that if I'm going to spend any time working on a plugin setup, I need a known working plugin first. It's like the chicken and the egg, and I spent an inordinate amount of time last summer, and earlier this year, chasing down bugs that were intermittent and apparently untraceable - because I was looking in the wrong places for things, like looking at the plugin manager when it was a plugin bug and vice versa.

This way, I can build it to not require any edits, and gradually convert it to make use of the newer add-on manager, whereupon I know it's working as it should be, without having to fix a bajillion things and not really knowing what's to blame and what's broken.

Plus it gives me a good way to apply my theory of how add-ons should work in a real and practical way; there aren't many plugins as big and complicated as SimpleDesk is, that integrate as tightly (but cleanly) as SD does to SMF, and believe me when I say it does an insane number of things, some of which I don't think have been done before or since.

For example, it actually modifies the language editor code, so that when you see the list of language files to edit, SD's are at the top and with 'nice' names for each file, not the mashup filename you normally get (and that happened because I was fed up dumping language files into the main languages folder and wanted to keep SD's language files to itself, in a very definite move that echoes how I want Wedge's add-ons to work)


As for the sidebar, I'm going to move all the SD profile sidebar items into the regular sidebar, and be done with it, so the problem goes away.

Re the name, I'm honestly not sure. WedgeDesk seemed natural but I don't worry about saying it many times fast - SimpleDesk pretty much fails that test too.

Mind you, the repo name is wedesk so maybe I'll end up switching it to that... Don't know, not really bothered by it at this stage...

Well, the hooks part was pretty easy to change over, the DB stuff was almost entirely automated, I'm currently working through the templates to convert them to Wedge coding but that's a bit tedious, and I'm also rewriting a number of things as I go, like making use of weToggle instead of half-baked hacky versions of the same thing.

Getting it all to play nicely is why I wanted to do it this way, so I know when I start using the proper add-on manager, that I make absolutely sure I'm hunting down add-on manager bugs and not WedgeDesk ones.

Right now there is still a fair bit that's fragile, but it is coming together and the parts I have tweaked (like the screenshot, putting the relationships and notifications blocks into the sidebar) look the better for it. I'm debating dropping some of the existing layout and heading towards a bit like the normal topic display does, with the poster info on the right... but we'll see, that's not finalised yet.

Oh, it's definitely going to be WedgeDesk, I already bought the domain names for that. I event spent some time mulling over how I think that site should look and function, seeing that it'll be just me designing for it.

In other news, I already have a list of things that I'm considering for future versions, which includes bug fixes of things I've found. (It's currently at 43 items, though some of those are notes for me for things to fix in the conversion)

I'm not averse to suggestions but it will very firmly go at my pace since there's no-one 'encouraging' me to work on this, so right now I'm concentrating on getting it up to speed for Wedge so that I can begin properly on the add-on and permissions overhauls. After that, I'll be more open to suggestions and so on, but the grand plan is to actually run WD as the primary module on wedgedesk.com, to function as both support vehicle and bug tracker. I mean, if you're going to do it, you gotta go for it properly, right?[1]

But I really want to have it not only out there to be a good mod for Wedge, to showcase what it can do - but also it'll be another way of promoting Wedge, and of promoting what can be achieved with sufficient care and attention. It may even help push Wedge's adoption not only by demonstrating that big, powerful add-ons can exist, but that there isn't the same issue against paid add-ons.[2]


@ Nao: It does use custom CSS. The enclosing div is not called either edge or edgehide, but it IS done by wrapping template layers around the entire content. When I said it was abuse of template layers, I really meant it: two divs are created, shd_left_column and shd_right_column. Left contains the left menu, the right column is opened in the _above layer, contains the main subtemplate, then closes it in the _below, which means it ends up containing the entire sidebar in it as well.[3]


I like Wedgets too, I actually nearly suggested Wedgelets quite early on but then I remembered how Java took that concept and screwed it up[4] with applets and servlets.[5] But Wedgets beats Wedgelets. And either of those are better than what I semi-seriously suggested for SD at one point.[6]

1.	The plan always was to move to SD on sd.net, since there's nothing technical really preventing it from replacing PT, it's mostly semantic difference, but a lot of that gap I can actually mitigate in WD with some of what I have planned.
2.	Even if paid mods end up being powered through wedge.org in some fashion, I'm still going to be running wd.com simply because that's the single best way to showcase it.
3.	I did not write the markup, and it's done that way because the menu reacts also to the << sidebar stuff in SMF, specifically that if the menus are dropdown, SD profile menu is a sidebar menu, and if the menus are sidebar, there's a little horizontal navbar above the content.
4.	Like most things in Java, really.
5.	Yes, servlets is a real Java term.
6.	"Drawers". On the basis that SD was a *desk* and desks have drawers. It made sense in my head.